Record for Spanglish variant
everson at evertype.com
Fri Jan 6 14:04:04 CET 2017
On 5 Jan 2017, at 19:42, Luc Pardon <lucp at skopos.be> wrote:
>> On Wed, Jan 4, 2017 at 5:38 PM Caoimhin O Donnaile wrote:
>> I wouldn't think so. Those have two clearly distinct languages in them, and should be recorded as such. Any detailed transcription would clearly separate the two languages at the word or phrase level, and have no problem doing so.
> I agree, and that is precisely one of the reasons why I don't think that an extensible mechanism for code-switching/hybrid languages is a good idea.
Exactly. There aren’t that many of them: certainly there’s no requirement to have Inuktitut/Thai, permitted by the (unreadable) extensible mechanisme. If we need a subtag for Latin/English and Nahuatl/Spanish, then we can have one.
> Sooner rather than later a well-meaning dictionary publisher will use it to tag his product, thereby (ab)using the mechanism for things it was not intended for.
> What happens here is that we expect that publisher - or any given tagger - to study the intended meaning of our extension in depth, and then make a judgment whether his mixed language text meets the criteria for defining an "hybrid" that we had in mind when we designed a mechanism for tagging them.
> I think that is expecting too much, and that therefore he should not be left alone to make that judgment. This kind of judgments, which require intimate knowledge of the various RFC's as well as linguistic considerations, is precisely what this list was supposed to make.
> Michael has demonstrated on this list that Spanglish is indeed sufficiently "special" to tell a Spanglish text apart from any other mixture of English and Spanish words. Once we are there, we may then as well assign a regular subtag for it.
> Extensible mechanisms are fine when no such judgment is involved, as is the case with the script subtag. It would be hard to make mistakes here, and therefore it is appropriate that a separate registration of - say - "ru-Latn" is not needed.
> However, hybrid languages are a totally different animal, and I think separate registration for each of them - and the oversight of this list that comes with it - is really needed to avoid tags spinning out of control.
> The number of hybrid languages like Spanglish may be very large,
It is A LOT smaller than the number of combinations that could be generated under an extensible mechanism.
> but that is a poor excuse to delegate our responsibility to the tagger. We should be perfectly capable to handle them one by one, as they come our way.
More information about the Ietf-languages